AP® Calculus AB & BC · Unit 4

How Does Linear Approximation Work in AP Calculus?

Linear approximation estimates f(x)f(x) near a point aa using the tangent line there: L(x)=f(a)+f′(a)(x−a)L(x) = f(a) + f'(a)(x-a). It works best when xx is close to aa, and whether it over- or underestimates the true value depends on concavity — concave up means the approximation underestimates, and concave down means it overestimates.

Linear (tangent-line) approximation

L(x)=f(a)+f′(a)(x−a)L(x) = f(a) + f'(a)(x-a)

Step-by-Step Method

  1. Choose a nearby point aa where f(a)f(a) and f′(a)f'(a) are easy to compute exactly.
  2. Write the tangent-line formula L(x)=f(a)+f′(a)(x−a)L(x) = f(a) + f'(a)(x-a).
  3. Substitute the xx-value you want to estimate into L(x)L(x).
  4. If asked whether the approximation over- or underestimates the true value, check the sign of f′′(a)f''(a): concave down (f′′<0f''<0) means the tangent line lies above the curve (overestimate); concave up (f′′>0f''>0) means it lies below (underestimate).
  5. Remember the approximation gets less accurate the farther xx is from aa.

Worked Examples

Foundational — estimating a square rootUse linear approximation to estimate 4.1\sqrt{4.1}, based on f(x)=xf(x)=\sqrt{x} at a=4a=4.
  1. f(4)=2f(4)=2, and f′(x)=12xf'(x)=\dfrac{1}{2\sqrt{x}}, so f′(4)=14f'(4)=\dfrac14.
  2. L(x)=2+14(x−4)L(x) = 2+\dfrac14(x-4), so L(4.1)=2+14(0.1)=2.025L(4.1) = 2+\dfrac14(0.1) = 2.025.
  3. Check against the true value: 4.1≈2.02485\sqrt{4.1}\approx 2.02485. The approximation is off by only about 0.000150.00015 — small, exactly as expected this close to a=4a=4.
Harder — over/underestimate confirmed by concavityUse linear approximation to estimate cos⁡(0.1)\cos(0.1) based on f(x)=cos⁡xf(x)=\cos x at a=0a=0, and determine whether it over- or underestimates the true value.
  1. f(0)=1f(0)=1 and f′(x)=−sin⁡xf'(x)=-\sin x, so f′(0)=0f'(0)=0 — the tangent line is horizontal here. L(x)=1+0(x−0)=1L(x)=1+0(x-0)=1 for every xx near 00, so L(0.1)=1L(0.1)=1.
  2. Check concavity: f′′(x)=−cos⁡xf''(x)=-\cos x, so f′′(0)=−1<0f''(0)=-1<0 — concave down, meaning the tangent line lies above the curve, so the approximation should overestimate.
  3. Confirm with the true value: cos⁡(0.1)≈0.99500\cos(0.1)\approx 0.99500, which is indeed less than L(0.1)=1L(0.1)=1 — the overestimate is confirmed exactly, not just predicted.

Common Mistakes on the AP Exam

How This Appears on the AP Exam

Try a timed AP-style question on this topic — free.

Start Free

Frequently Asked Questions

What is the formula for linear approximation?

It is L(x)=f(a)+f′(a)(x−a)L(x) = f(a) + f'(a)(x-a) — the value of the tangent line to ff at x=ax=a, evaluated at the point you want to estimate.

How accurate is a linear approximation?

It is most accurate close to the point aa, and the error grows as xx moves farther from aa — it is only ever an estimate, never guaranteed to be exact except when ff really is linear.

How do you know whether a linear approximation overestimates or underestimates the true value?

Check the concavity using f′′(a)f''(a): if the function is concave down (f′′<0f''<0) near aa, the tangent line lies above the curve and the approximation overestimates; if concave up (f′′>0f''>0), it underestimates.

What is the difference between a linear approximation and a differential?

They use the same formula in different notation: the differential dy=f′(a) dxdy=f'(a)\,dx describes the CHANGE in the approximation, while L(x)=f(a)+f′(a)(x−a)L(x)=f(a)+f'(a)(x-a) gives the approximated VALUE itself — L(x)=f(a)+dyL(x) = f(a)+dy.

Is linear approximation tested on the AP Calculus AB exam?

Yes. Linear approximation is core Unit 4 content tested on both the AB and BC exams.

Related Topics

Keep going with timed tests & full mocks, study materials for every unit, DME Nova — your AI tutorthat won't make up an answer, and FRQ feedback scored against an exam-style rubric.

Start Free

Written by the DME AP Calculus team — 20+ years of AP & STEM teaching.

Last updated: 2026-10-08

AP® is a registered trademark of the College Board, which is not affiliated with, and does not endorse, this site.